![]() | ||
|
|
Water Block Design / Construction Building your own block? Need info on designing one? Heres where to do it |
![]() |
Thread Tools |
![]() |
#1 |
Cooling Neophyte
Join Date: Jan 2004
Location: Sweden
Posts: 17
|
![]()
Well, just (allmost) finished my first CPU block. I designed it in solidworks2003 änd made it with mill down to 1/100 mm precision. Mutcho hard work!
So, yesterday I put on the top, silver soldered it. The base i 12 mm thick and the top 60 mm. What happened? Yes, the block got to hot and "bent", or "screwed". The thing is that the bottom of the channels are quit thin, approx 1.2 mm.... So if I lap it down I probably will end upp having zero thickness.... Anyone having an idea? Can i make the "middle" of the block flat anyway? Only over the core that is.. |
![]() |
![]() |
![]() |
#2 |
Cooling Savant
Join Date: Feb 2003
Location: Willmar MN/Fargo ND
Posts: 504
|
![]()
I supose if you stuck it back in the milling machine and faced it down. you could fix that.
Do you have any pictures/information/links about this block you made to 1/100mm precision? thats some tight tollerances I would say. How much is it bowing? I dont see how it could bow too much if at all, I havent had problems like that from soldering. Maybe you could machine another and make a screw on top. 12mm and 60mm? thats really thick! Jon |
![]() |
![]() |
![]() |
#3 | |
Cooling Neophyte
Join Date: Jan 2004
Location: Sweden
Posts: 17
|
![]() Quote:
Pics? Well i forgot to take pictures before I mounted the top, but i have SolidWorks pics of it. And well, the mill is calibreted to 1/100 mm, and I set the values exact... I actually didn't have time yesterday to check the bowinbg exact, I'll do it today, but perhapes 1 - 2 mm. And yes, make an other, but the milling process alone tock approx 10h, and I don't have the time right now...I just hope I can fix the old one.... Sending pics when I can(working right now....) |
|
![]() |
![]() |
![]() |
#4 |
Cooling Savant
Join Date: Feb 2004
Location: W. Sussex, UK
Posts: 329
|
![]()
heat it up and bend it back? your soldering was probably with too much heat, or for too long to make it bend...
|
![]() |
![]() |
![]() |
#5 |
Cooling Savant
Join Date: Sep 2002
Location: Sweden
Posts: 318
|
![]()
Listen to me: you can never bend something back! its pointless, trust me. What you can do is either mill it down or lapp it down...or worse, throw it away :/
|
![]() |
![]() |
![]() |
#6 |
Cooling Savant
Join Date: Feb 2004
Location: W. Sussex, UK
Posts: 329
|
![]()
it was just a suggetion... I havent seen a pic of it, how badly it is bent, so anything suggested may not be helpfull, we just dont know enougth about it.
If you anneal it to make it soft again, then let it cool, then you can hit it back into shape gently with a hammer. you probably wont be able to if your design has made the copper weak though. this will not make it perfect either, it will still need to be lapped down slightly, this may just give you less to lap, and give you a thicker base once its done |
![]() |
![]() |
![]() |
#7 |
Cooling Savant
Join Date: Mar 2003
Location: St. Paul, Minnesota, USA
Posts: 123
|
![]()
The best you can probably hope to achieve by some kind of straightening is to get it closer to flat so that there is less to correct.
I think that as long as the block isn't horridly out of shape, you should be able to flatten and lap the area that touches the core without having to worry about the rest of it. If you do this, be extra careful when mounting because the distance between the base and the pad rests on the CPU (if you're using a CPU that has them) will probably be off. You don't want to crack a core just because the pads hit the block funny.
__________________
Abit NF7-S ... xp1800+ ... Abit GF4ti4200 ... Maxtor 120gb SATA150 Currently gathering parts to water cool everything! (read about the project) |
![]() |
![]() |
![]() |
#8 |
Cooling Savant
Join Date: Nov 2003
Location: Suffolk, UK
Posts: 234
|
![]()
Everything will warp when you solder it. Use a fly cutter in the mill or lap it
Ramember it only has to be flat over the cpu die. It would have to be pretty badly warped to move 1.2mm over a 10-20 mm area. |
![]() |
![]() |
![]() |
#9 | |
Responsible for 2%
of all the posts here. Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
|
![]() Quote:
The heat on one side expands the metal, and if the heat isn't uniform, the block can bend. So heat the other side, and let it bend back. You'll still need to do a lap job on it, but you can concentrate it on the CPU core area, to minimize it. Interesting... I guess I'll hit the same problem when I put together my HDD block. |
|
![]() |
![]() |
![]() |
#10 |
Cooling Savant
Join Date: Feb 2004
Location: W. Sussex, UK
Posts: 329
|
![]()
I have never noticed any changes when I have heated copper or silver. I know they will change slightly with heat but I have always assumed it would be only a very small amount not worth noting. Today my dad took the newly made copper lid for my gpu block to work so he could silver solder it, which would need about 2/3x higher soldering temp than normal solder. When he gets home I will measure it and compare it to the previous measurements I made. Im hoping that it hasnt changed much
![]() Edit - It hadnt warped atall, not noticably anyway. It still fits the outline i drew around it perfectly... Last edited by |kbn|; 02-25-2004 at 06:15 PM. |
![]() |
![]() |
![]() |
#11 |
Thermophile
Join Date: Sep 2002
Location: Charlotte, NC
Posts: 1,014
|
![]()
Have you considered putting in a vice to see if you can press it flat? Also, is it concave or convex above the core?
__________________
I have a nice computer. |
![]() |
![]() |
![]() |
#12 |
Cooling Neophyte
Join Date: Jan 2004
Location: Sweden
Posts: 17
|
![]()
Thanks for all the sugestions!
But the eproblem is solved! The bottom was thick enought so I could lap it down. It is not perfect, two corners are still a little "bent", but thats extreamly little... The thing is, the bottom plate is now only 0.5 - 1 mm thick. ![]() BTW The block is a pinblock, drilled and milled. Holes with vertical channels. The pinns have that turbulence creating shape, still they are quit big since they are not only created by milling holes...Pics will come...or solidwork pics... |
![]() |
![]() |
![]() |
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|